P. Huy, “Beyond the AI feedback paradox: How metacognitive re-flection transforms AI corrections into L2 learning”, RSELTL, vol. 4, no. 4, pp. 823–846, Jul. 2026.